Gmail activates artificial intelligence feature for spam emails

Gmail's latest update introduces AI-powered spam detection to protect users from unwanted junk mail. According to the search giant, this update not only improves user security but also prevents spam.
 Gmail activates artificial intelligence feature for spam emails
READING NOW Gmail activates artificial intelligence feature for spam emails
Gmail’s latest update introduces AI-powered spam detection to protect users from unwanted junk mail. According to the search giant, this update not only improves user security but also significantly improves the spam detection rate.

Gmail is now more secure against spam emails

The world’s most popular e-mail service has implemented its most important update in recent years. Gmail, which aims to defeat spam senders through advanced artificial intelligence, now has a new text classification system called RETVec (Resilient & Efficient Text Vectorizer). This new spam detection technology will work against text manipulation, a common tactic used by spammers to bypass traditional filters. Spammers often use special characters, emojis and other techniques to evade spam protection.

These manipulations often involve homoglyphs, which are characters that resemble standard alphabet letters, making them difficult for traditional spam filters to detect. Developed by Google, RETVec overcomes this challenge. Rather than relying solely on character recognition, the new feature uses a 200,000-parameter machine learning model that uses visual similarity.

The new approach significantly improves Gmail’s ability to block spam and phishing attempts by enabling it to understand and filter emails that use deceptive tactics. According to Google, the introduction of RETVec increased spam detection rates by 38% and reduced the error rate by 20%. This means users are less likely to see spam in their inboxes and less likely to miss important, genuine emails that are mistakenly marked as spam.

Comments
Leave a Comment

Details
154 read
okunma47757
0 comments